ABSTRACT Vibration‐based systems have shown that appropriately oriented and tuned mechanical oscillations can disrupt regolith interlocking and promote flow. Building on this evidence, the present study investigates vibration‐controlled granular discharge as a potential ISRU (In Situ Resource Utilization) strategy, with the aim of supporting predictable material processing through physically consistent numerical modeling. The equations governing the dynamics of nonspherical particles are presented accordingly and used to simulate the behavior of a fixed mass of granular material flowing through a hopper device in the framework of the discrete element method (DEM). The results show that even for monodisperse distributions vibration‐induced fluidization does not follow simple monotonic trends. The mass flow rate initially increases with the forcing frequency f , reaching a plateau in the range of 150–200 Hz, beyond which further frequency increases produce no significant enhancement. At the same time, higher values of the dimensionless acceleration Γ generally reduce the mass flow rate and cause an increase in the frequency threshold above which vibration‐driven discharge exceeds gravity‐driven flow. Moreover, for fixed, moderately low values of Γ, increasing f leads to the emergence of higher‐order harmonics in the instantaneous mass flow rate. For fixed and relatively high values of f , a larger value of Γ can cause the appearance of subharmonics in the spectrum. In contrast, for relatively low f and high Γ, incommensurate spectral components appear, which reduce the regularity and predictability of the flow response. These findings are supported by a detailed analysis of particle patterning behavior and frequency spectra, which reveals the fine‐scale dynamics underlying the observed flow regimes.
ABSTRACT Granular assemblies are prone to form self‐organized dissipative structures. The microscopic interparticle sliding, entailing frictional dissipation at contacts drives configurational changes that critically influence the mechanical stability of the system. Using the discrete element method (DEM), this study investigates the mechanical stability of granular assemblies based on a mesoscale second‐order work criterion. In dense specimen, the vanishing of the mesoscopic second‐order work exhibits a spatial localization after the stress peak, consistent with the formation of a shear band. By analyzing the evolution of stability indexes, the sub‐domain inside the shear band undergoes a fluctuating regime of stability, which governs the fluctuating stability characteristic of dense specimen at the critical state. For the loose specimen, this fluctuating stability regime maintains throughout loading up to the critical state. The shear band domain and the whole loose specimen were identified as inherent dissipative structures, governing the fluctuating stability regime, where the different loop categories composing the topology of the granular assembly share similar stability mechanisms. Furthermore, the dissipative structure exhibits a similar mean characteristic size of unstable meso‐loops at the critical state regime. The meso‐loops with sufficiently large instability influence range contribute to a loss of stability of the dissipative structures, ultimately controlling macroscopic instability of granular assembly.
ABSTRACT Blasting excavation can induce permanent displacement in faulted rock slopes and adversely affect slope stability. Most existing stability assessments use deterministic approaches and therefore do not adequately account for the spatial variability and cross‐correlation of geomechanical properties. In addition, conventional assumptions of transverse anisotropy and the arbitrary selection of copula models may bias blast‐design assessments. This study develops a stochastic framework that integrates rotated anisotropic random fields with copula models to characterize spatially variable rock properties. The effects of copula selection and rotated anisotropy on the statistical distributions of permanent displacement and a displacement‐based risk index are investigated. The framework is demonstrated through an engineering case study and compared with conventional deterministic analysis. The results indicate that the commonly used Gaussian copula tends to overestimate slope stability, whereas strata orientation, scale of fluctuation, and the correlation between cohesion and friction angle significantly affect the response statistics. The case study illustrates the applicability of the framework to comparative stability assessments of blast‐affected faulted rock slopes.
ABSTRACT Long‐term stability of rock engineering structures is controlled by progressive strength degradation under sustained loading, yet direct experimental determination of long‐term strength (LTS) is severely constrained by impractically long testing durations and strong material variability. To address this limitation, this study proposes a feasible numerical framework for long‐term stability analysis by linking time‐dependent microscale damage evolution to engineering‐scale strength parameters. The four‐dimensional lattice spring model (4D‐LSM), incorporating creep, plasticity, and progressive bond fracture, is calibrated using conventional creep test data, in which the maximum bond‐fracture threshold governs long‐term failure time. Time effects are introduced through temporal evolution of model input parameters, enabling numerical creep tests to reproduce stress‐time‐to‐failure behavior and predict LTS. The numerically obtained LTS data are fitted with an empirical stress‐time relationship and further converted into time‐dependent equivalent Mohr‐Coulomb cohesion and friction angle. These time‐varying parameters are then implemented in a classical strength reduction finite element framework, and a machine learning model is adopted to establish the correlation between the factor of safety and the degrading strength parameters. The proposed approach facilitates the practical prediction of long‐term stability of rock engineering structures using temporally evolving strength parameters derived from 4D‐LSM simulations, thereby providing a mechanistically sound and computationally efficient tool for the assessment of delayed failure in rock engineering.
ABSTRACT Experimental laboratory investigations underlie most of the knowledge about the mechanics of soils under non‐isothermal conditions and their thermally induced deformation. Despite advances, the understanding of thermally induced deformation of soils remains inconsistent, hindering the development of a coherent framework that can advance science and engineering. Considering this challenge, this paper presents an analysis of the established methodology to determine thermally induced soil deformations from laboratory experiments. Established equations are re‐derived, verified, and compared, and the resulting predictions are analyzed. Data support that the quantification of thermally induced soil deformations can be affected by volume corrections and calibrations, the integration of conservation equations under non‐isothermal conditions, the misuse of thermal expansion coefficients, and the lack of repeatability tests and uncertainty quantification. Quantitative analyses support the conclusion that these aspects contribute to the contradictions in the state‐of‐the‐art. Therefore, this work proposes an updated methodology to experimentally quantify the thermally induced deformation of soils, whose widespread adoption is supported by the provision of an open‐source toolbox that implements the proposed changes to the state‐of‐the‐art: The Thermally Induced Deformation Analysis Library, TIDAL.
ABSTRACT Floor water inrush threatens the safe mining of the mine. The automatic roadway formation without pillar by roof cutting (ARFPRC) method can reduce the floor water inrush risk. However, the key parameters for mining above confined aquifers are difficult to determine. In response, this study proposes a design method for determining the key parameters of ARFPRC based on floor failure. First, the unique advantages of ARFPRC in reducing floor water inrush risk is elucidated. Then, numerical simulation studies are conducted to analyze the characteristics of floor damage under different roof cutting parameters, leading to the determination of the optimal parameters. It is also clarified that grouting reinforcement to improve rock properties is beneficial for preventing water inrush. Subsequently, a field test of floor grouting reinforcement was conducted on site. After its effectiveness was confirmed through geophysical detection, the ARFPRC industrial test was carried out. Finally, based on the measured depth of floor failure, it was demonstrated that the designed parameters effectively controlled floor damage. The research findings provide a theoretical foundation and engineering guidance for the application of ARFPRC technology in mining areas with the risk of floor water inrush.
ABSTRACT The operational safety of existing shield tunnels is increasingly threatened by construction activities in the proximity domain. Accurate prediction of their longitudinal response is thus of great importance. Previous research usually adopted the Winkler or Vlasov foundation models coupled with Euler‐Bernoulli beams to establish analytical solutions. However, the solutions were tailored solely to and available for a specific disturbance scenario, which required an extension when encountering diverse ones. In addition, they are insufficient to capture the behavior of the soil continuum. To address these issues, this paper presents a comprehensive model by integrating the more advanced three‐parameter Kerr foundation model with Timoshenko beam theory. By introducing the state‐space method, a unified solution procedure is developed for predicting the tunnel response. The method's validity and versatility are demonstrated through comparisons with field measurements and existing solutions under three distinct scenarios: new tunnel undercrossing, adjacent pit excavation, and surface surcharge. Finally, a sensitivity analysis for the excavation case examines the influences of soil elastic modulus, pit‐tunnel angle, and burial depth. The proposed framework offers a unified and efficient approach applicable to various disturbances, providing an efficient tool for the prediction of shield tunnel deformation.
ABSTRACT This study presents a fully coupled hydro‐mechanical implementation of the Clay and Sand Model (CASM) extended to partially saturated conditions (U‐CASM) within the COMSOL Multiphysics environment. The formulation adopts Bishop's effective stress and suction as stress variables and incorporates alternative loading–collapse surfaces, dilatancy laws, and a porosity‐dependent water retention curve. The main contribution lies in a fully customised implementation strategy based on symbolic algebra, which enables user‐defined constitutive modelling, simplifies equation integration, and reduces programming effort within a multiphysics framework. The model is verified against benchmark simulations for clays and sands under drained and undrained conditions in both saturated and unsaturated states. Validation against triaxial and oedometer data confirms the ability of the model to reproduce soil behaviour over a wide range of stress paths and suction levels, while highlighting the influence of alternative constitutive assumptions. A field‐scale application further demonstrates the capability of the framework to capture key hydro‐mechanical features of partially saturated soils, including collapse upon wetting.
ABSTRACT Coupled numerical models based on CFD and DEM were developed to analyze the interaction between fluid flow and sediment particles. To enhance simulation efficiency and accuracy, a series of methods were proposed, validated, and applied, including the single‐phase oscillation method, the coarse‐grained method, the porous medium replacement strategy, and a semi‐resolved approach. Numerical simulations were performed to investigate the scour process around a submarine pipeline situated on a slope, specifically examining the effects of slope angle, pipeline suspension height, and wave height. The flow field, forces acting on particles, and particle movement patterns were analyzed. Results indicate that while the slope angle significantly affects the flow pattern, it exerts limited influence on hydrodynamic forces. The decomposition of gravity into components alters the force balance on particles, rendering them more vulnerable to erosion and facilitating downslope transport under wave action.
ABSTRACT The paper investigates the impact of two classical relative permeability formulations, one expressed in terms of pore water pressure and the other in terms of degree of saturation, on the predicted seepage in an infinite slope subjected to surface pore pressure variations. A hysteretic retention law is employed, inducing irreversible changes in hydraulic conductivity during wetting–drying cycles. Two soils with different gradings are analysed and, for each of them, the two relative permeability formulations are matched at a reference point so as to represent the same material. Despite this calibration, the corresponding permeability fields differ significantly over the hysteretic domain, by as much as five orders of magnitude. These discrepancies are mostly the consequence of retention hysteresis and are substantially reduced if a unique, non‐hysteretic retention curve is selected. For reference, a slope model with constant permeability, equal to the maximum saturated value, is also evaluated. The magnitude and rate of the slope response to surface perturbations, the number of cycles to restore equilibrium and the characteristics of the steady state regime all depend on the chosen permeability formulation. Slopes with a constant permeability react rapidly and uniformly, whereas slopes with permeability depending on pore water pressure or degree of saturation display slower, more complex and depth‐attenuated responses, particularly for the finer soil. Overall, the results highlight the need for consistent calibration of both relative permeability and retention hysteresis, while underscoring the limitations of constant permeability models in capturing unsaturated seepage, even when appropriate hysteretic retention laws are adopted.
ABSTRACT The EAC‐1A lunar mare simulant, a silty sand, was selected by the European Astronaut Center (EAC) for use in the LUNA analogue facility's lunar simulation testbed. Its stress‐ and strain‐dependent shear modulus and damping ratio were determined via resonant column testing on specimens prepared either from ambient moisture material or oven‐dried material. Significant differences were observed across the strain range. The tests further examined specimen behaviour during unloading from large to small strains and subsequent reloading. The reloading curves were observed to merge with the virgin response curves at the strain corresponding to load reversal, while a notable degradation of the small‐strain modulus due to prior loading was evident. The overall behaviour in terms of shear modulus and damping was captured using simple equations.
This study proposes a novel geometrically regularized gradient-damage model for simulating dynamic mixed-mode fracture in orthotropic materials with tension-compression asymmetry. In this model, a thermodynamic framework is formulated by incorporating damage dissipation into internal energy evolution, from which the constitutive relation and the damage energy release rate are derived. The geometry of sharp cracks is regularized using a functional of crack surface density, resulting in a volumetric expression for the Griffith-type crack dissipation energy. By enforcing energetic equivalence between the crack dissipation energy and the damage dissipation energy, a gradient damage evolution law is obtained. An orthotropic Helmholtz free energy decomposed by an orthogonality-based volumetric-deviatoric-spectral split operator is introduced to consider the material anisotropy and tension-compression asymmetry. A hybrid driving force for unified modelling of mixed-mode fracture, is then proposed by combining the decomposed free energy with the Mohr-Coulomb criterion and three mode-dependent fracture energies. The resultant governing equations are discretized within the finite element framework and solved via an alternate minimization Newton-Raphson algorithm. Its accuracy and robustness are verified through three benchmark problems of 2D and 3D dynamic fracture. The results demonstrate the proposed method is powerful in modelling complex dynamic anisotropic mixed-mode fracture.
Accurate stope stability assessment is essential for safe underground mining. This study develops an interpretable, data-driven framework to update the conventional stability graph and reduce its reliance on predefined linear boundaries. Based on 426 cases, a multilayer perceptron optimized by balancing composite motion optimization achieves a test accuracy of 90.70%. Stratified bootstrap resampling confirms stable lower bound performance, with 95% confidence intervals of [0.85, 0.95] for accuracy, [0.83, 0.95] for balanced accuracy, and [0.85, 0.95] for weighted F1 score. Monte Carlo perturbation of rock mass and geometric inputs produces only about a 1% label-flip rate, indicating robustness to moderate input uncertainty. Sensitivity and interpretability analyses identify the stability number as the dominant factor, while hydraulic radius acts as a secondary variable with strong interaction effects. Residual Kriging is then used to calibrate predicted probabilities in the stability graph, select an optimal decision threshold of 0.3, and define an uncertainty zone for ambiguous conditions. Relative to existing stable-unstable prediction boundaries, the updated graph shows stronger class separation, and external validation on two independent datasets yields conservative accuracies above 90%. The framework is mainly intended for open stoping and open-span stability assessment; application to mining methods governed by different failure mechanisms requires site-specific recalibration and validation. Overall, the study provides an interpretable and uncertainty-aware pathway for quantitative stability graph refinement.
Intergranular strain extensions of hypoplasticity (IGS, ISA, GIS) are widely used to model small-strain and cyclic soil behaviour. However, recent studies and our own simulations reveal that these models can exhibit stress overshooting not only under repeated small loading cycles but also during monotonic loading. This manifests as an unphysical accumulation of deviatoric stress and mobilised friction angles exceeding , leading to overpredicted shear strength and stiffness. Of the three inspected extensions, IGS and ISA are the most susceptible to overshooting. While GIS mitigates overshooting under deviatoric loading, it persists during oedometric unloading-reloading. The cause lies in the constitutive structure of intergranular strain frameworks: when the mobilisation variable satisfies , the effective elastic strain range becomes too large relative to the prevailing stress state, allowing stress accumulation with insufficient plastic dissipation. To regularise this behaviour while retaining the hypoplastic backbone, a two-step projection approach is proposed that activates only upon violation: (i) a deviatoric projection onto an auxiliary Matsuoka-Nakai surface with a state-dependent cut-off angle; and (ii) an isotropic projection that limits the mean effective stress to a state-dependent bounding pressure consistent with hypoplastic barotropy. Validation across four benchmark problems, including triaxial compression tests with monotonic loading and cyclic reloading, oedometric compression with unloading/reloading cycles and a laterally loaded pile, demonstrates that the projection is non-intrusive when the stress state remains admissible, and effectively regularises overshooting when it does not. The approach improves model robustness and physical consistency in simulations involving small-strain cycling or expanding plastic zones.
Reliable design of shallow foundations is critical for offshore and coastal infrastructure such as towers, tanks, and wind energy structures. This study presents a framework that combines finite element limit analysis (FELA), random field theory, and ensemble machine learning to evaluate the probability of failure (PF) of ring footings in spatially variable soils. Random fields of soil friction angle (phi) and unit weight (gamma) are modelled with lognormal distributions and exponential autocorrelation, while Monte Carlo simulations with FELA generate a comprehensive reliability dataset under varying coefficients of variation, correlation lengths, and safety factors. Several surrogates, including support vector regression, multivariate adaptive regression splines, artificial neural networks, and Kolmogorov-Arnold networks, are benchmarked. A stacked model integrating KAN and ANN with a support vector meta-learner achieves superior accuracy (R 2 = 0.993) and consistent calibration. Parametric analyses show that PF is mainly controlled by safety factor and soil variability, while correlation length and ring slenderness exert secondary influences. The framework further provides reliability design charts that incorporate geotechnical variability and geometric effects. These results highlight the efficiency of surrogate-assisted random field FELA for reliability-based design, offering offshore engineers practical tools to assess foundation safety under uncertainty.
This study introduces an approach relying on the application of an operational matrix based on shifted Chebyshev polynomials for numerically addressing fractal-fractional (FF) linear and nonlinear differential equations, as well as systems of equations, utilizing the generalized fractional derivative of Caputo type. The proposed method transforms the generalized Caputo-type FF derivatives transformed into a system of algebraic equations, enabling the determination of unknown solutions. Theoretical analysis was conducted to establish convergence criteria and derive error bounds for the method. The approach was validated through quantitative analysis across various scenarios and benchmarked against established techniques to affirm its precision and computational effectiveness. Additionally, the method was applied to the SIRD (Susceptible-Infected-Recovered-Deceased) mathematical model with a FF operator, employing the spectral collocation method to demonstrate the effectiveness of the proposed numerical approach in handling FF derivatives.
T-shaped pressure cast-situ-solidified soil pile with a spray-expanded frustum (T-PSPSF) is composed of fluidized solidified soil, and features distinct upper and lower pile sections with varying diameters, an expanded body with double-frustum (EBDF), and ribbed plates. T-PSPSF exhibits superior performance including high bearing capacity, enhanced construction productivity, and notable economic and environmental benefits. Based on the laboratory test results, the mechanical parameters of the pile are clarified. The geometric configuration of the pile was determined through field tests. For practical purposes, the diameter prediction method of EBDF is proposed. Additionally, numerical simulations were conducted with varying diameters, heights, and positions of EBDF, as well as different ratios of upper pile diameter to lower pile diameter, to determine the optimal pile structure. The load-sharing ratio of the EBDF to the total pile head load exceeds 50%. Considering both the bearing characteristics and economic efficiency of the T-PSPSF, the optimal diameter of EBDF is adopted as 1.4D, the optimal length is 1.0D, the optimal position can be adopted between 5D and 7D above the pile head, and the optimal ratio of upper pile diameter to lower pile diameter is taken as 0.5.
ABSTRACT Multilayer perceptron (MLP) networks are predominantly used to develop data‐driven constitutive models for granular materials. They offer a compelling alternative to traditional physics‐based constitutive models in predicting non‐linear responses of these materials, for example, elastoplasticity, under various loading conditions. To attain the necessary accuracy, MLPs often need to be sufficiently deep or wide, owing to the curse of dimensionality inherent in these problems. To overcome this limitation, we present an elastoplasticity informed Chebyshev‐based Kolmogorov–Arnold network (EPi‐cKAN) in this study. This architecture leverages the benefits of KANs and augmented Chebyshev polynomials, as well as integrates physical principles within both the network structure and the loss function. The primary objective of EPi‐cKAN is to provide an accurate and generalizable function approximation for non‐linear stress‐strain relationships, using fewer parameters compared to standard MLPs. To evaluate the efficiency, accuracy, and generalization capabilities of EPi‐cKAN in modeling complex elastoplastic behavior, we initially compare its performance with other cKAN‐based models, which include purely data‐driven parallel and serial architectures. Furthermore, to differentiate EPi‐cKAN's distinct performance, we also compare it against purely data‐driven and physics‐informed MLP‐based methods. Lastly, we test EPi‐cKAN's ability to predict blind strain‐controlled loading paths that extend beyond the training data distribution to gauge its generalization and predictive capabilities. EPi‐cKAN achieves superior accuracy in predicting stress components and generalizes well under blind strain‐controlled loading paths. It maintains robustness to noise, achieving only 1.52% error in deviatoric stress predictions with 5% noisy data, outperforming MLP models.
ABSTRACT A numerical investigation is performed on one‐dimensional coupled heat transfer and solute transport in saturated, incompressible layered soils, incorporating the effects of three soil phases: solid particles, mobile pore fluid, and immobile pore fluid. The finite difference method is employed to solve the coupled governing equations for heat transfer and solute transport. The thermal module of the model assumes local thermal equilibrium between solid and fluid phases, and incorporates heat transfer through conduction, advection, and mechanical dispersion. The solute transport module accounts for diffusion, advection, mechanical dispersion, and sorption processes. The proposed model is first developed and subsequently subjected to verification. Numerical simulations demonstrate that heat transfer considerably influences solute transport, affecting key metrics including breakthrough time, mass flux, cumulative mass outflow, and concentration distribution within the soil stratum. These effects persist not only during active heat transfer but also long after its cessation. Furthermore, effective porosity significantly impacts both heat and solute transport processes, and the Soret coefficient also notably affects solute transport through thermodiffusion. Conventional solute transport analyses that neglect the coupled effects of heat transfer and effective porosity may yield substantially divergent results.
